GOCE引力梯度的频谱分析及滤波

Frequency analysis and filtering processing of gravity gradients data from GOCE

GOCE卫星提供的梯度数据含有非常大的低频误差,如何处理这种误差是GOCE数据处理中最为关键的工作之一.本文根据GOCE卫星的运行情况,首先分析了梯度数据的频率特性,推导了频率与阶次的对应关系;并在此之上,介绍了针对低频误差的滤波方法,即移去恢复和向前向后滤波方法,前者可解决滤波中的低频信号损失问题,后者则主要解决了滤波中的相位漂移问题.最终结果表明:引力梯度的时间频谱与球谐展开中的阶次虽不是一一对应的,但各阶所对应的最大截止频率与阶次却有一定的显式表达.同时也表明,本文所采用的滤波方法是有效的,达到了消除低频误差但保留观测频段信号的目的.

How to remove the low frequency error is a very critical job for GOCE data processing,since the error is so huge.This paper firstly analyzes the frequency characteristics of the gradients data in time domain,and derives the formula which can show the relationship between the degree and frequency;then two main filtering strategies which are remove-restore method and forward-backward filtering technique will be introduced.The former can solve the problem of low frequency signal loss in filtering processing and the latter can solve the phase drift problem.The final result shows that the formula given by this paper can show the relationship between the maximum cut-off frequency and degree of spherical harmonic model correctly.It also shows that the filtering method given by this paper is effective and can remove low frequency error but save signal in measurement band width.